About IBD Podcast Episode 194 — How to Protect Yourself in the Age of AI

How to Protect Yourself in the Age of AI — About IBD Podcast Episode 194

Leave a reply

Artificial intelligence (AI) is here to stay. How can we approach AI with both curiosity and caution, especially as it becomes more integrated into daily life? If we’re not aware of it and looking for it, AI could become inserted into spaces where we aren’t comfortable with it. Amber talks with her son Brenden, a computer science student, about recognizing AI-generated content, the importance of media literacy, and the ethical challenges technology brings. This conversation offers some practical uses for AI, such as learning new skills or automating mundane tasks, while pointing out that creativity and critical thinking are human pursuits. We stress how AI isn’t a replacement for expertise and that we should all question what we see online to encourage the responsible use of AI tools.

Listen and Subscribe to About IBD: Apple Podcasts | Spotify | Overcast | Castbox | Amazon Music | All Other Podcast Apps

Find Brenden Tresca on:

Find Amber J Tresca at:

Find Mac Cooney (mix, sound design, and theme music) at:

These show notes may contain affiliate links. If you choose to purchase after clicking a link, Mal and Tal Enterprises, LLC may receive a commission at no extra cost to you.

Transcription

[Music: IBD Dance Party]

Amber Tresca 00:00:05  I’m Amber and this is about IBD. I was diagnosed with ulcerative colitis as a teen and had pouch surgery ten years later. Now I help people with Crohn’s disease and ulcerative colitis understand their disease and feel seen on About IBD. You’ll get real talk on symptoms, treatment, and life with IBD straight from patients, caregivers and experts. You’ll learn, feel less alone, and even have a laugh along the way.

In this episode, we’re going to talk about how to approach AI with both curiosity and caution. We’ll talk about how to spot misinformation and disinformation, and how patients can vet what they’re seeing before trusting it. Plus, we’ll go over a few ways people can use AI tools that isn’t a replacement for expert advice.

My guest is my son, Brenden Tresca. Brenden is a college student majoring in computer science. He has a real interest in artificial intelligence and how it fits into everyday life. During school breaks, Brendan works as an IT professional, which gives him a practical, real-world perspective on how people actually use technology.

Amber Tresca 00:01:17  He also hosts a radio show called Tech Fest at the University of New Haven, where he discusses emerging tech and digital trends. And somewhere in the midst of all of that, he’s an Eagle Scout and a third-degree black belt in taekwondo. Brenden, welcome to About IBD.

Brenden Tresca 00:01:41  Thank you so much for having me. I’m very happy to be here.

Amber Tresca 00:01:44  So this is, I believe, your second appearance on About IBD. The first appearance may have been 2020 or 2021. I’m not exactly sure when we did that. I didn’t look it up, but I may have to put in a clip because you were much younger then.

Brenden Tresca 00:02:04  Much younger? Yes. Much younger, I do remember.

Amber Tresca 00:02:08  Brenden, let’s start with your interest in AI. Would you tell me a little bit about your focus on AI and how you were looking at it, both professionally and personally?

Brenden Tresca 00:02:20 Well, that’s a great question. I think that I try to look at AI because I believe it is really sort of the future, and I know you can sort of get that sort of response from, you know, every, every third tech pro.

Brenden Tresca 00:02:34  But I really do think that AI, you know, whether we like it or not, is going to integrate itself into all parts of our life. You know, be it personal, like you said, professional. all the work that we’re doing, especially learning, I think with like, again, whether you like it or not, AI is kind of here to stay. And that’s why I think it’s worth looking at. Because if I don’t look at it now. I’ll be. I’ll have to look at it later. So that’s that’s really why I take such an interest in it. Because I really do think it is part of our future.

Amber Tresca 00:03:05  Yeah. For sure. Brenden, do you consider yourself a tech pro?

Brenden Tresca 00:03:09  I’ve been called one way too many times for me to completely deny it, but. Oh, boy, I was just just putting that out there. Yeah.

Amber Tresca 00:03:18  That’s funny. All right, how about for yourself? Why are you using AI? Or are you?

Brenden Tresca 00:03:28  Every once in a while, yes.

Brenden Tresca 00:03:29  When there’s a tricky question that, you know, I can. I look at the internet. Reddit won’t help me. I will go to the AI because I even though I understand I’ve almost, I would say almost actually preached the fact that it cannot be completely accurate. It is not to be trusted. And it is something, a tool to be held with caution. I also do believe that it can give me more personal answers and it’s just easier to understand.

Amber Tresca 00:03:56  Right. I feel like in the beginning, when I started using AI and I used AI to write your bio for this podcast just to see what I could get out of it. So I gave it your LinkedIn profile and asked it to write a bio. It did not do a great job. It missed out on a couple of really key points in your LinkedIn profile. Even though I told it that I needed this bio for a podcast episode on AI, it really wanted to focus on your status as a third-degree black belt. So I had to have it circle back and say, that’s not the focus of this.

Amber Tresca 00:04:41  The focus is how he is engaging with tech and his background there. So rewrite the bio with that in mind. And I probably went through four iterations before it gave me something workable. Even then, I went back and edited it further. So my point with this is that I still needed to have a certain amount of knowledge about you and about what I wanted, because the AI could not get me there. I still had to edit it, I still had to direct it. And so I’m wondering, is that because AI just wasn’t able to understand me? Like, do I need to learn better how to train it, or is it because it’s never going to quite get me to? It’s never going to quite give you exactly what you’re looking for, because that’s just not the nature of it.

Brenden Tresca 00:05:43  I think probably the latter is what you should expect. It’s not going to. AI, especially ChatGPT, has been trained to have a very sort of and it’s hard to explain, but mostly sort of like a, like a very, like a very nice, a servant that just loves to serve.

Brenden Tresca 00:06:00  And it’s always, well, what else can I do for you? Is that helpful. And it’ll, it’ll really focus on some ideas while also cut out others. And depending on who you are and just the answers and questions that you are putting into this machine. I really do think that it’s not necessarily your fault the AI is going to talk a certain way, and the people that ask it are going to talk in all sorts of different ways. It needs to be edited, it needs to be reworked, it needs to be looked at. It’s not just something that you could say, all right, write me a podcast episode. All right, all right: pretend to be my son so you could be on my podcast. Like, it’s not going to. Exactly. You need to do you need to do a little bit of tweaking to it. It can’t exactly give you what you’re looking for. So you did the right thing, is what I’m saying right.

Amber Tresca 00:06:46  Yeah. And to your point, I had to have some knowledge, not only of you, but also of how the tool worked in order to ask it.

Amber Tresca 00:06:59  And I think I had to ask it four times before I got what I actually wanted. And the thing about that is I’m wondering if it actually saved me time. I’m not sure because I’m an experienced writer, I know you. I probably could have sat down and wrote your quick bio, but I just wanted to go through that exercise to see what it would give me. I will say some years ago I asked it to write your bio, and it hallucinated a lot of things because you didn’t have a digital footprint, then you were not in, you know, over the age of 18 as you are now. And so it just decided to tell me a whole bunch of things that were completely fake.

Brenden Tresca 00:07:40  Yeah, it likes to do that.

Amber Tresca 00:07:41  Sometimes I don’t say please and thank you to it anymore.

Brenden Tresca 00:07:47  It’s really just a formality at this point.

Amber Tresca 00:07:50  Because I was just typing things in the way that I would type an email, for instance, that I would type an email and I would say, would you mind or would you consider or would you please send me x, y, z? And then at the end I would say, thank you so much.

Amber Tresca 00:08:08  I basically was doing that in the beginning because that’s just normally how I communicate with a human. And so then I really took it in that I’m not communicating with a human. I’m communicating with a machine. Do you think that matters? Do we need to be communicating with it that way, or is it right to be thinking of it differently and changing our communication style?

Brenden Tresca 00:08:38  I think that really depends on what you’re asking for. If you’re, I know the I looked up the usage of AI, how people are actually using ChatGPT And, I found out that roughly, I want to say and don’t quote me on this, but roughly 50% of people were using it for some sort of like, role play aspect where it was supposed to play a character. You know, they were they were I, I don’t ask about the specifics of that, but the main point is the, material that you’re asking for can vary wildly depending on if you’re saying please and thank you or, you know, slamming the papers down on the desk and saying, get this done by three.

Brenden Tresca 00:09:17  And, honestly, even though that everyone is looking for something different from AI and it’s going to give you sometimes the same outputs, I do think that at this point, saying please and thank you is not going to get you a drastically better answer than it was if you were just just to be direct and say. Baa baa baa baa baa baa. Because, I personally believe that if you’re just trying to get a product. You don’t need to be conversational. I mean, when you need, you do it to a human because a nice human is going to give you better things than a mad one. So, I would say that it really depends on the person, but personally, I don’t think it’s necessary. I really don’t.

Amber Tresca 00:10:04  Okay, that’s good to know because I’ve kind of given up on that. And I am rather more like, get this done by 3 p.m., than I am a, you know, please and thank you at this point. So, Brenden, I don’t really ascribe to this idea that older people can’t engage with technology in the same way that young people do.

Amber Tresca 00:10:27  Although I will admit to needing to find a millennial on occasion to explain things like Snapchat to me. So. But once I get a little training, I’m good. But I’m wondering about your thoughts on that and if there’s anything that people your age and I’m thinking kind of specifically like being in school, because I think being in school is a lot different than being in the workforce. Are there things that people your age understand about AI that other generations might not have caught up with yet?

Brenden Tresca 00:11:06  Yes, I do think so. The general understanding about how maybe not AI works, but sort of the outputs of AI, and our ability to look at something and say that right there that was clearly ChatGPT-generated, or that was that’s not a real video, that’s not a real image. And, you know, I, I hate to push a stereotype, but I feel like one of the ones that comes up the most is older people on Facebook sharing AI videos of cats or dogs or like cute little toddlers.

Brenden Tresca 00:11:37  Right? This is obviously fake. This is obviously AI generated and everyone’s commenting like, oh, such cute babies. Oh, such a nice story. Oh, you know, whatever. And I think, I wouldn’t want to chalk it up to media literacy, because I feel like that’s too slippery of a slope, and we have our own problems on that, where I think people who are younger will investigate and not be as ready to consume content so willingly with such a cavalier ish attitude, attitude where, you know, I don’t care what it is about. It’s a cute cat and I’m done with it.

Amber Tresca 00:12:11  Right? And so this brings me to the idea of you don’t know what you don’t know. And do you think that people your age who may be using AI in school, or maybe learning about it in school are more aware of the existence? And so the idea that something could be AI generated, quote unquote fake occurs to you where might not occur to somebody who is not using AI, or is not studying it, or is not as familiar, do you? Do you think that that plays a role?

Brenden Tresca 00:12:53  So I, I do definitely think that it is our usage of AI that sharpens our senses as to what is AI-generated.

Brenden Tresca 00:13:02  You know, someone who may not be utilizing a technology or utilizing to it to an extent where it’s new and needs to be learned as much as we do, that can lead to them not having as great of a sense for what it is. I mean, I think that appears for any profession, I’m sure, like you see you if you do not do all sorts of audio work all the time, you might hear bad audio in the wild and be like, wow, that sounds good to me. But now that you’ve done it so much, you have an ear for it and now you can understand. You’re like, I know what you’re doing. I know that’s wrong. I know, I know this needs to be made. And I feel that like for, video editing and movie editing and all sorts of stuff like that. Before, I didn’t have a care in the world about what it was. Everything was great to me. But now that once I did it, I can see you did this trick.

Brenden Tresca 00:13:54  You failed to do this. You just get a sense for it. And I think that’s really going on with AI.

Amber Tresca 00:14:00  Yes. You make an excellent point. I think it’s also very funny that you pointed out how I work in audio, and now I have a higher bar for the audio quality of things, and I may not be as willing to listen to something with poor audio or podcasts that are recorded over zoom, quite frankly, because I can tell that that’s what they did, whereas another person might not care and might tell me to, get over myself.

[MUSIC: About IBD Transition]

Amber Tresca 00:14:36  Up next, the ethics of AI and how that affects the ways that we’re using these tools.

Amber Tresca 00:14:52  In one of your classes this past semester was the ethics of Technology. Can you tell me about some of the discussions that you had in that class with other students and with your professor, and how that is a integral part of your education as a computer science major?

Brenden Tresca 00:15:15  Well, that was honestly, in my opinion, one of the one of the most exciting classes I’ve ever taken.

Brenden Tresca 00:15:22  It was it challenged everybody’s ideas on AI, technology, ethics of things in general that we previously had thought of as two taboo or uncomfortable to even think about. And it was through our professor’s way of teaching, which is a very diplomatic way to put it, that we had our ideas and our previously thought of, prejudices and whatnot challenged and brought out to a point that every if you were not uncomfortable, you were not learning. It really made me think about how AI and technology is. It has its own biases. It is being used inappropriately. It will be used inappropriately. There are certain, it was sort of that sort of, idea that it’s like if you give something to the public, the worst possible thing that could happen to it will happen to it. That’s that that was something that we learned. And that’s obviously what’s going on right now with everything publicly available. whatever is the worst possible outcome, it is already happening. It is already being used. And, if I would like to be a software developer, I would have to think about and I would have to make usage cases for every aspect of my program and how it could be used, whereas as a user, I would just use it in the way I want, and I wouldn’t think about how people are using it for awful, horrible things.

Brenden Tresca 00:16:57  I really do think it’s something everybody should think about and that, you know, you are your privacy, your ability to do things. Your in this case, your media literacy is everything is in trouble and everything is challenged. And you really should be thinking about this from a more, critical standpoint rather than just everything’s fun. I use AI to generate pictures of cats, and I don’t have to worry about any of the problems. There are problems, and you may be the cause of them. So that’s something to keep in mind. And that’s definitely a very important thing in computer science.

Amber Tresca 00:17:35  So that’s a little dark.

Brenden Tresca 00:17:39  I know that’s that’s the that’s the not so good part.

Amber Tresca 00:17:44  I, I don’t disagree with that because we have seen time and time again that, something will be developed. Whatever it is, AI programming, social media. And if the developers fail to anticipate how real humans are going to interact with it and use it, that can lead to loopholes that people are going to use for…

Amber Tresca 00:18:20  Maybe as we as parents usually say to children for making poor choices. So I feel like maybe that’s what your class was getting at, was instilling in all of you that are going on to be computer science professionals, that you need to anticipate how the end user might use your tools, not for what you designed it for, but maybe for something else.

Brenden Tresca 00:18:51  Exactly.

Brenden Tresca 00:18:53  and that that I think in still having a dark side of it and basically instill a little bit of fear mongering, I’m not gonna lie, helps you to, not think everything’s death darkness, but, balance out your…what we all previously had conceived as, like, a happy, fun awesomeness of technology. It’s, you know, it’s pushing that scale to now be even, you know, all this, all this bad stuff versus all this good stuff. Now we have an even sense of how to use technology, how to make technology, all that sort of stuff.

Amber Tresca 00:19:32  Right. And to your point about the fake videos, for instance, one of the downsides being that it is getting us used to seeing something that’s created by AI and not necessarily differentiating, and then also maybe not even caring. Whether it is or not.

Brenden Tresca 00:19:58  It’s well, an AI generated video of a kitten, I guess that’s cute enough. It just keeps getting worse and worse. And as you get used to it, it opens you up to getting that little bit, that small amount of uncomfortable being uncomfortable, which then is smoothed out by, I guess it’s okay, and then it gets a little worse, but I guess it’s okay. And then you get bad things, so.

Amber Tresca 00:20:22  Right, right. Yeah. I feel like maybe that is the issue there. Do you think it’s different if Something were watermarked, or there was some other identification that says, okay, this is a funny little video of a cat, and you’re going to laugh at it or go or whatever. But just so you know, it’s not it’s not real. Like it’s okay. I mean, obviously like we watch animated movies and things like that, right? Like, but the thing about an animated movie is, is that, you know, you know, that someone, you obviously drew it.

Amber Tresca 00:20:55  It’s very clear, you know, that, it’s not always the same with AI. So do you think there’s a responsibility there for either the AI platforms or the people that are using them to identify them that way, and not to try to pass them off as something that’s real.

Brenden Tresca 00:21:13  Is definitely on the companies that are making these tools to, have some sort of safeguard in place a watermark, a, you know, audio jungle in the middle of your stolen audio.

Amber Tresca 00:21:28  Yeah, I forgot about that.

Brenden Tresca 00:21:30  That always got me. That always got me. Yeah, but something  like that. Even though now any tool, any audio tool could take that audio jungle out in seconds. It’s a difficult thing to think about because there really is no way to stop content being stolen reposted. I mean, that’s what social media is all about. Half of the stuff is not stuff that was made by the original person who’s posting it. It is being repurposed, reused, edited, crunched down, recycled into new content.

Brenden Tresca 00:22:00  And it would be better if there was a logo in the first place. Because even though we know for a fact the end result is no logo, it’s stolen. People think it’s real. That also means that the first person who posted it, there’s going to be a slightly less chance of that. Being believed is real because there is a watermark, but, you know, still down the line, it’s not going to make a difference. However, you would rather have one than not have one.

Amber Tresca 00:22:26  We know that we can’t prevent certain things, right? Like we all lock our doors, right? We all lock our car doors, we lock our front doors. If someone really wants to get into your car, wants to get into your house, they’re going to do it.

Brenden Tresca 00:22:39  They can do it.

Amber Tresca 00:22:39  But yeah, but you still lock the doors. So I feel like that’s kind of what we’re talking about here. And there might be people that say, we don’t need to put a watermark on something, or we don’t need to clearly identify it as being AI generated because people are just going to cut the watermark off, or people are just going to repurpose it anyway, and then it’s going to it’s going to get lost.

Amber Tresca 00:23:03  So why should we even start? Why should we even try? I feel like that’s where we are. But there are people that say that type of thing.

Brenden Tresca 00:23:12  Yeah. And I think that’s totally not true. And like you said, if you wanted, if  I was going to break into your car and it was locked, I would probably just notice it was locked and then run away. But in the range of people that are going to bring in, break into your car. A Swat team can do it anyway, but if you locked your car, that means the lowest common denominator. Probably me trying to break into your car is not going to be successful. So even though you know out of if you rank it from 1 to 10, how dangerous the person is that’s trying to break into your car. You would rather be safe than sorry is the easiest way to put it right?

Amber Tresca 00:23:50  Right. That’s a good point. It’s going to be a deterrent for some, you know, people who are technologically advanced are going to be able to do things, but that’s going to be a smaller proportion of the population.

Brenden Tresca 00:24:06  Exactly.

Amber Tresca 00:24:07  So. Okay. Good point. Okay. So we’re talking about how people can create things using AI. Right now there’s no I mean, I know some of the social media platforms have a little button or a toggle that you’re supposed to use to identify something as AI generated. But I don’t think that there’s any, any penalty or any risk if you don’t identify it as such. So this stuff is already out there. We have to have a high level of skepticism for the things that we’re looking at. I think when it comes to cat videos, maybe not such a big deal. When it comes to health data. I think it is a much larger deal. So I’m hoping that you can help me and help our listeners understand what can we look for, what can we think about so that we can differentiate something that has been created by AI? Are there tools people can use? Is there something like, just to the naked eye that they can look at it and identify it? Like, how can we help folks figure out something that’s AI generated versus something that’s not?

Brenden Tresca 00:25:24  That’s a really good question. I’m actually very glad you asked. so that one of the biggest things that I go by when I’m trying to and this can go, I say this first because this this can anybody can do this. You don’t need to have a special eye. You don’t need to be, you know, thinking very quickly, you don’t have to really be there. Logic. Right? I saw a, I saw a video that a friend sent me yesterday of a cat, and it was a cat going through a plastic tube. And maybe on the first two watches, I was like, wow, what a cute kitty cat. Then I realized from the perspective of the camera, you can see there were like multiple shelves or whatever, and you can realize that the camera, like, goes through those shelves to watch this cat. And then I realized this actually isn’t real. But if it were not for that specific piece of logic, like thinking about it, oh, it’s it’s not really. Everything else looks very realistic.

Brenden Tresca 00:26:22  But logically, does that make sense? How could the cameraman have seen this if there are shelves in the way and he’s not having to move back around the shelves, he’s just going through them, right? So number one would be logic. Does it make sense? Would this really actually happen in the first place if it’s too ridiculous to be true? It’s probably too ridiculous to be true. the other thing I would say is especially in AI videos, and not a lot of people know about this look for noise. Noise is sort of like when you see a lot of pixels change on the screen. You might see them like fizzle out a bit before they, return to, to to like, staying normal, you know, like I it’s really hard to explain in words, but just sort of the way that pixels will change color or they may it may look like things are like shimmering a little bit before they settle down. That’s a sign of AI videos. AI videos as well as AI images are generated from pure noise, and the algorithm then sorts that noise to be like, okay, what can I build from this? What can I build from what I just built and so on and so forth.

Brenden Tresca 00:27:31  So that noise that you see in the video is the sign of AI, because when a lot of pixels have to be changed, the AI algorithm can exactly keep up as well. So some of that noise still does still escape. And I like thank God for that because that’s such a crazy like, whistle blow in the AI videos because they otherwise look very real. Another thing I would look for, and this has become very scarce and it’s probably this won’t be a good piece of advice to go by anymore within 1 or 2 years. But, look for things that don’t line up. You see, you can see all these things behind me. You can see how they are parallel to each other. There are no tangent lines there. They’re not intersecting each other. And I might merge these two corners right here. It might. It might change up these little lines. It might change the details of these characters. And you won’t notice unless you get really close. And then you can see that, well, you know this.

Brenden Tresca 00:28:38  His hand and his arms seem to be, like, connected right here. And when they should be connected at the elbow. So, that’s another thing. And the, the third thing I would look for is especially with, especially with videos is voices. Listen, listen to it a billion times and you will notice that it will sound like this. Do not talk to them that way. Right. That doesn’t sound human. As if I were say, don’t talk to them that way. You notice tone. Notice maybe there’s artifacts in the in the voice. Maybe it’s sounding like. That’s right. That doesn’t sound right. Basically, my end goal is look at every aspect of it and you can’t really know it’s AI until you basically sort of consume it a little and then realize, okay, this is AI, I’m not going to share it. I’m not going to like it. I’m not going to do this. I’m not going to pass it off to other people as real. Because now I know now that I’ve looked at it over and over and over again, I can tell it’s not real.

Amber Tresca 00:29:41  Right? Thank you for this, because there’s a few things that I didn’t know either.

Brenden Tresca 00:29:46  It’s uncomfortable unfortunately.

Amber Tresca 00:29:49  Yeah, yeah. What about text-based things? Because I see these also coming out. I think I might know a few tells. Hopefully you might know a few more. One of the things with AI generated, for instance, Instagram captions or even a longer form, you know, blog posts, things like that. One of my biggest tells is if there is a big change in someone’s content, for instance, like you can you can get a sense of and as a writer, I think maybe I pick up on the nuances a little bit more, but I can tell someone’s writing style, someone’s writing voice, and then one day it shifts. And then I wonder, are they now utilizing an AI tool in order to create this content? That’s not necessarily a bad thing, but I do wonder to your point about the ethics of doing that and then not disclosing it. So aside from that, do you have any other ideas about how you could identify a written post or a blog that might have been just wholly generated by AI?

Brenden Tresca 00:31:10  well, there’s almost so many topics to think about, I can’t even list them all.

Brenden Tresca 00:31:17  One like I said, is logic. Does this really make sense. Is this. Would someone really write about this. but especially for things that are ChatGPT generated or just AI generated text. one thing, I think you already know this em dashes. That’s that’s a very I know, like it’s a real.

Amber Tresca 00:31:33  Thing, not stop using em, dash.

Brenden Tresca 00:31:35  I know I won’t either.

Brenden Tresca 00:31:36  And every time I do it, I’m like, wait, I have to really make sure that this doesn’t make me seem less human. But aside from that, a another one that I see all the time. ChatGPT won’t be interesting. It’ll be awesome. You see, you see, it was blank. Won’t be ex. Yeah. It’s why it’s it’s. And that’s always paired with the em dash is, it’s always like this bringing up a small counterclaim and then refuting it immediately. Most of the time when you’re talking about something, it’s odd, and especially when you’re talking about something that doesn’t need that level of like, gravitas to it.

Brenden Tresca 00:32:13  Why would you say that unless an AI generated for you who doesn’t have a sense of context, who doesn’t care.

Amber Tresca 00:32:19  So are you saying like over the top? Like because I’ve noticed that in in using tools, you know, writing tools that they will like, I will use words that I know that everybody can understand. I’m writing so that people can…I’m writing for humans who have all different levels of comprehension when they’re reading. So but then AI will say, no, you should replace this word, you know, with another word. So I will say, if you’re going to eat fiber XYZ and it will say no, no, no, you should say for people who consume fiber. And I’m like, nobody. Like, that’s not helpful. Yeah, that’s not how.

Brenden Tresca 00:33:02  That was going to be my next point is, if I’m just talking about picking apples, ChatGPT will be like, this wondrous man has picked the best, flavorful fruits in the entire, city of New York. They will surely be bring greatness, greatness to everyone.

Brenden Tresca 00:33:20  No one talks like that, right? It’s it always tries to bring this like this very fairy like amazingness to everything. And the truth is, in the real world, people don’t care. People don’t like, love things to that extent. They especially that’s not the way they’re going to talk about them. They’re going to say, I picked apples today. It puts importance in places where you really don’t need it. It’s needlessly just pumped full of this, like Hawthorne level of detail, which is not necessary at all.

Amber Tresca 00:33:54  And I I’m dying because you are my child. That you just described something as Hawthorne level.

Brenden Tresca 00:34:01  Yes, yes, I, I just can’t.

Amber Tresca 00:34:05  It’s hilarious.

Brenden Tresca 00:34:05  I but his his work was something that I felt the same way it was it was like ChatGPT was a real man. He was talking about things that I got it. You don’t need to tell me how beautiful the bushes were. You don’t need to tell me how hard that door slammed and how angry everybody looked. I get it right.

Brenden Tresca 00:34:26  And it seems like you were just. He was just filling up word counts the whole time. So that would be the biggest tell in my opinion.

Amber Tresca 00:34:33  That’s hilarious. I actually kind of loved that. Although I remember in my AP English in in high school, I was the only one that that enjoyed that kind of writing. Everyone else was like, yeah, it’s going on forever about nothing. And I was like, yes, but so many cool words I just liked. I just like words. So but that’s a really good point. Being too verbose and being too excited about something that’s maybe mundane.

Brenden Tresca 00:35:04  Exactly, exactly.

[MUSIC: About IBD Transition]

Amber Tresca 00:35:10  Coming up, the fine line between having AI teach you to do something versus having it do it for you.

Amber Tresca 00:35:27  What is something that you hope that our listeners today don’t do with AI? After hearing this episode and remember, let’s keep this PG or G, let’s keep it in that in that range. We’re going to leave all that other stuff off, right?

Brenden Tresca 00:35:45  Yeah. I don’t need to speak about that.

Amber Tresca 00:35:47  No.

Brenden Tresca 00:35:48  No, I think, I guess this is more of a personal gripe. People always do all sorts of, like, little fun fantasy, like role plays with it. I wouldn’t suggest that. I think that’s a that takes a huge toll on your mental health, talking to robots all day. But, I would say do not have things do tasks for you do not have things. Take away the creative aspect of whatever you’re doing. Do not have AI…don’t make…don’t make things where there was no point in having you ask. If you’re asking ChatGPT to do all of this, why do I need you in the first place? If you’re going to that level, there’s no point of you being in the in the equation at all. So I would suggest that when you’re making I mean, I make I make everything I try to make games, I make art, I make music, I would never oh especially code. I do not get ChatGPT to write my code for me.

Brenden Tresca 00:36:48  I say, what’s going on here? What’s wrong? What could I change? Because otherwise no one needs me as a programmer. They could just ask ChatGPT if that’s what we if that’s what we were doing. So, do not let AI be creative because it really can’t be by a technical standpoint, right?

Amber Tresca 00:37:07  And so it’s probably a good time to explain how I use AI in the creation of this podcast. I don’t have an AI trained, but I do have an AI tool that will, for instance, break a podcast episode into chapters. So it will tell. It will give me like timestamps on the beginning of a new a new subject. Usually it’s like with, you know, every question starts a new topic. It will also do a machine generated transcript for me. It will also do things like suggest a title, that type of thing. Now it’s using the podcast episode that I give it. I’m not asking it to make a podcast episode. I’m asking it to help me think about the main points of the episode, and then also to transcribe it so that I can make sure that the content is accessible, no matter how people want to learn.

Amber Tresca 00:38:09  So I actually don’t have any concerns about using AI in that way. I think it’s helping me be more effective in helping make my show available to more people, and doing things like breaking it into chapters helps people find what they’re looking for. So I think it’s definitely a value add, because if I did not have the tool assisting me in this way, those things wouldn’t get done. I guess is what I’m saying. So I’m not using it in a creative way, which by the way, to date I have yet to have an experience with any of the AI tools that I’ve tried help me in a creative way. They just can’t. I also use I use Canva to create many of the images that I, I use in the social media assets that I create for this show. They do have something in there that is supposed to be able to generate things or help you. I have yet to successfully use it for anything, so I’ve rather given up on it because I am better off working with a designer who actually can design something based on what they learned in school and through their career, versus asking the AI.

Amber Tresca 00:39:31  It just can’t do that for me yet.

Brenden Tresca 00:39:33  Yes. that’s that’s basically one of the best uses for AI. In this case, you are doing something where your main part of the work is still overshadowing how much the AI is doing. If I want to listen to a show, and I found out that you generated the whole thing with ChatGPT, then I don’t need to deal with you at all. I can just generate my own. And I think letting ChatGPT fully take over all of what you’re doing is not the way to go. There has to be human ingenuity and imperfection to it, really, because otherwise there’s no point in you doing it at all. Then it’s not your creation. You didn’t make anything. Unfortunately. So. Right.

Amber Tresca 00:40:19  There’s a podcast that I listened to that I actually had a concern at a certain point in listening to it. And it is because to your point about imperfections, I wasn’t hearing like those slight imperfections that you usually get with, with audio medium. And I thought to myself, I wonder if they used an AI tool to create the voice, at least the voice, the writing, you know, I don’t know, but the voice and I had such a visceral reaction to it and I was like, if this is AI-generated, I don’t want to listen to it. I kind of was surprised at my own reaction. I didn’t really expect to react that way. Right? So I went back and looked and engaged with this, show creator, more thoroughly and discovered, no, they do not. It is not AI. It is actually this person’s voice. And then I felt better about consuming the content. Right. But I was really like, I’m not going to listen to this anymore. If it’s AI-generated. And it kind of surprised me that I felt that way because to your point, like, like, would it really matter? It’s just like a history podcast. It wasn’t conversational in any way. So I guess it just goes to show how I feel about AI-generated content in a certain way.

Brenden Tresca 00:41:47  That’s probably, that’s a good reaction, I believe, to have just be like, no, I do not want this. I’m not listening to a human’s work. But yes.

Amber Tresca 00:41:56  Yeah. All right. So on the other side of this coin, is there anything that you would encourage people to use AI for or something that they should…that they should try and see if it works out for them.

Brenden Tresca 00:42:11  Yes. And that is learning. I’ve done it with math. I’ve done it with coding. I’ve done it with all sorts of things where I don’t know what to do. I don’t ask the AI to do it for me. I ask, how do I do this? And once it tells me the instructions, I can do it myself. And learning things like language or code or really anything is, I think, one of the best things you can get out of an AI, because it is, it is it’s getting close to doing the creative work, but it’s actually the output is you. Right? You are the output. Your knowledge and what you understand about a subject is becoming the output. And there you can use the AI.

Brenden Tresca 00:42:56  Because of the AI, you can do things more efficiently as a human. And I think that’s the best. That’s like the penultimate, right? That is the best way to use AI in my opinion, is getting other people to to learn. You can ask it how to do a math equation, and instead of having it do the math equation for you, which lowers your knowledge because you’re not learning, it tells you how to do it, and then you can do it on your own. And it’s like having a teacher there for you. And that I would highly encourage people to do is ask it to teach you things, ask it to help you learn, ask it to do whatever. Because especially with that often two conversational tone that comes in really handy because it feels like not really talking to a person, but basically like talking to a person more than just like a robot, which will tell you step one, step two, step three, step four. Do it right. I can’t help you if you don’t know what’s going on.

Brenden Tresca 00:43:50  I don’t know, that’s all. That’s that’s all she wrote. So, that’s definitely what I would encourage people to do is learn use AI to help you learn things.

Amber Tresca 00:43:59  That’s a really good point. And I think it’s a fine line between asking it to do something for you and asking it to teach you how to do the thing.

Brenden Tresca 00:44:08  Right.

Amber Tresca 00:44:09  And math. Your example of math, I think, is a is a really valid one. Yeah. It can. It can. It can take you through the steps on how to solve the equation. And then you’ll learn how to do it yourself versus just having it.

Brenden Tresca 00:44:22  Do the answer.

Amber Tresca 00:44:23  Do it for you. Just give me the answer. Yeah. Yeah. I feel like, Brenden, we’ve been a little. Well, we’ve been really critical of AI, which maybe that’s important at this stage because it has sort of come on the scene. And for right now, there’s no there’s no training. There’s no, you know, learning to drive for these things.

Amber Tresca 00:44:51  They’ve just been introduced into the wild. And people are playing with them and doing things with them. And yet, perhaps none of us are fully cognizant of how they’re going to impact us yet. which almost makes us feel a little bit like we’re the guinea pigs. And that as things shake out, there’s going to be some of us who are not going to be as happy with how we’ve used AI. So I would like to end though on maybe we can find some positive things, maybe we because I don’t think that either you or I looks at AI as completely negative. We don’t want to get rid of it. No, I think we’re both worried about giving people the tools to understand what it is and how it works, and how we can use it to better humanity versus as replacing, for instance, creativity. So, I don’t know. Do you have any final thoughts on the on the benefits of AI? And for instance, is there anything that you’re looking forward to with AI, you know, as these tools improve maybe?

Brenden Tresca 00:46:10  I think things that humans don’t want to do in the first place is something that can be done by by AI, you know, like writing long paragraphs about things that no one really, really cares about or, you know, factory work or something like that. Even though it seems like you are taking away a human job, I believe it would be better to do something that people really don’t want to do. Then, you know, take something like art or music or coding, which is that’s all about what people want to do. You only do that because you know you really like it. So, I, I do really think that the benefits of AI lies in it, replacing the things that we never wanted to do in the first place. I didn’t want to sit and shovel coal all day. An AI-powered shovel could do that for me. And then we’re all happy because then I can spend my life doing more creative things, right? Like experiencing things. I can live my life as a human, whereas the AI doesn’t care.

Brenden Tresca 00:47:13  It will just do the work. It has no feelings. So that’s that’s what I think we could do with it, is make sort of these not to fizzle out humans in the workforce, but basically make workers that don’t need to the the bad parts about being a human don’t apply, which is I don’t want to do this job. Someone that will do the job means that the human that would have done that job can go do better jobs, right can can go do more interesting jobs, jobs that require less mundane work that are more, challenging for the mind that leave them happier, basically. And I think the short of it is getting AI to do things for us like clean the house or wash the dishes can improve our our quality of life.

Amber Tresca 00:48:02  I agree with you 100%. I would absolutely, yeah, I don’t need AI to record podcasts for me, but I would love AI to do the effing laundry and the effing dishes. and there’s something that I see a lot of people and this is going to be polarizing.

Amber Tresca 00:48:20  And so I would expect someone to take issue with how I’m looking at this. The, automated checkout systems that are in almost every major retailer. Now, there are folks that don’t like those because they their concern is, is that that’s taking a cashier job away from a human. I personally love using the self checkouts most of the time for various reasons. And my thought on it was that what if then at the grocery store, the grocery store then has the, the budget and the funding to invest in putting a dietitian in the store, for instance. What if the grocery store had a dietitian that you or, or me or anybody could go and talk to and could work with and they could go through the store with us and help us meal plan things like that. That’s something that we need a human to do. That’s something that you want to have a conversation with somebody about. What are your allergies? What are your food triggers? What do you like to eat? Help me figure out how to make a more healthful meal for my family.

Amber Tresca 00:49:42  And this person could help you. Meanwhile, the actual scanning of the items. So you know how much to pay the store could be something that is done by a robot, right? So, I mean, that’s my thought on that. And I know, it’s a it’s a polarizing thing, but I’m hoping that that’s something that we can get to with AI is that we’re looking towards how it can benefit us. I don’t need it to write music. I, I need it to help me with the things that are preventing me from writing music. And I say that like, I’m not writing music, but you know what I’m saying? Yes.

Brenden Tresca 00:50:28  Yes. I wholeheartedly agree.

Amber Tresca 00:50:30  Brenden, I have one more question for you, and I didn’t put this in the episode outline. Inside baseball, we do work from an outline so that we make sure that we’re covering things. I’m wondering, Brenden, if Peter Griffin lived with an IBD, what would he say?

Brenden Tresca 00:50:52  [Peter Griffin imitation] Wow, this fricking thing sucks.

Brenden Tresca 00:50:57  I, I probably wouldn’t have a lot of, words that I could say on the show here to talk about the matter. It would probably shock it. It would probably become a experience in the future where he would say. [Peter Griffin imitation] This is worse than the time I had IBD.

Brenden Tresca 00:51:13  So. Yeah, that’s, Thank you for giving me that opportunity to do that stupid voice.

Amber Tresca 00:51:22  I had to get. I had to get one of your impressions in here somewhere, because they are. They. Oh. Well. Okay. All right, I thought I was going to bring something out of left field for you. Brenden, do you want to talk about your radio show for a minute and where folks can find it and listen to it?

Brenden Tresca 00:51:42  Oh, yeah. Of course. so I am Brenden Kenny, “the jackhammer” of Tech Fest on 88.7 WNHU. The official New Haven student run radio. Tech Fest, talks about everything tech within the week, what’s going on, what you need to look out for. It’s really just my rundown, my thoughts and my advice on what the hell is going on right now. you can if you have a radio.

Brenden Tresca 00:52:09  If you have car, you can listen to it. 88.7. We are there. We’re there all the time. I’m there every single Friday from 7 to 8. So be there.

Amber Tresca 00:52:19  Sounds good. I’ll put more information in the show notes on where you can stream the radio show. Because if you’re not in the greater New Haven area, it might not come through on your radio if you even have one. so, Brenden, thank you so much for doing this with me. AI is coming at us from so many different directions, and I expect that this is going to be an ongoing conversation. I’m going to talk to a lot of different people regarding AI and how it’s being used in healthcare. So I appreciate your giving an overview of maybe a little higher level, how we can think about it, the ethics of it, and then how we can try to tell the difference between something that is human-generated versus machine-generated. And as I was thinking through this topic, I was suddenly struck that I actually have two experts in my household on AI, you and then also your dad, my husband, Mike, runs a newsletter on AI, and he is a very experienced user of AI, so I’m going to speak with him as well. So thanks for doing this with me. I hope that this brings you more listeners to your radio show because it is entertaining. You do occasionally break out an impression here and there so folks can get more of that on your on your radio show. And best of luck in your next semester studying computer science.

Brenden Tresca 00:53:54  Yes! Thank you so much. It’s been really nice being on the show.

Amber Tresca 00:54:00  Hey super listener! Check the show notes to find links and information about the topics discussed in this episode. Plus, get a written transcript and much more on my website About IBD. If you enjoyed this episode, please consider sharing it with someone else or leaving a rating in your podcast app. Reviews and ratings help me grow this show and bring you more great content.

Thanks for listening and remember, until next time, I want you to know more about IBD.

About IBD is a production of Mal and Tal Enterprises.

It is written, produced, and directed by me, Amber Tresca.

Mix and sound design is by Mac Cooney.

Theme music is from Cooney Studio. 

Leave a Reply

Your email address will not be published. Required fields are marked *